[Some problems of studying spatial heterogeneity of animal communities]

The efficiency of different techniques of data collection and processing was considered by the example of bird communities. Accounted variance of the coefficients of bird complex similarity was used as the measure. Annual censuses in different sites proved more expedient than a long-term census in a single key site to study medium- and small-scale spatial heterogeneity of bird communities. Averaging and generalization of data collected in different years and sites yielded satisfactory results. Estimated informativeness in this case proved closer to the real one compared to averaged long-term data since the included representation errors in space and time.